Sie sind auf Seite 1von 11

Downloaded from http://oaj.oxfordjournals.

org/ at UNAM on January 31, 2012

The Surviving Image:

Georges Didi-Huberman

Downloaded from http://oaj.oxfordjournals.org/ at UNAM on January 31, 2012

The Surviving Image: Aby Warburg and Tylorian Anthropology


Georges DicH-Huberman

1. Aby Warburg, 'Italian Art and International Astrology In the Palazzo Schifanoia' in The Renewal of Pagan Antiquity- Contributions to the Cultural History of the European Renaissance, trans. David Bntt (Getty Research Center for the History of Art and the Humanities. Los Angeles, 1999), p. S85. 2. See Georges Didi-Huberman, Derant l'image. Question posee auxjlns d'une histolrt de Van (Minuit: Paris, 1990) 3. See Aby Warburg, Gesammelte Sduiften, eds G. Bing and F. Rougemont (Teubner: Leipzig and Berlin, 1932). Significantly, the longejt entry in the very precise index of this edition (four pages, three columns) is devoted to the expression Nachleben. See also dirs Hans Meier, Richard Newald, and Edgar Wind, Kulturwissenschaftliche Bibhoaraphie xum Nachleben da Antike-A Bibliography on the Survival of the Classics, (Cassell: London, 1934), p. 5, where the near impossibility of translating this term into English is already noted. 4. See Aby Warburg, Schlangenntual. Em Reiseberlcht (1923), ed U. Raulff (Klaus Wagenbach: Berlin, 1988 and 1996); Fritz Saxl, 'Warburg's Visit to New Mexico' (1929-1930), Leaura (The Warburg Institute. London, 1957), pp. 32530; A Dal Lago, 'L'arcaico e il suo doppio', Aut out, no. 199-200, 1984, pp. 6 7 91; C. Naber, 'Pompeji in Neu-Mexico. Aby Warburgs amerikanische Reiie', Freibeuter 38, 1988, pp 88-97; P Burke, 'Aby Warburg as Historical Anthropologist', Aby Warburg. Alaen der intematlonalen Symposiums Hamburg 1990, dirs. H. Bredekamp, M. Diers, C. Schoell-Glass (VCH-Acta Humaniora: Weinheim, 1991), pp. 39-44; Kurt W. Forster, 'Die HamburgAmerika-Unie, oder. Warburgs Kulturwissenschaft zwischen den Konnnenten', Aby Warburg Alaen inurnaUonalen Symposiums Hamburg, pp. 1137 and 'Aby Warburg: His Study of Ritual on Two Continents,' October 77, 1996, pp. 5-24; S. Settis, 'Kunstgeschichte als vergleichende Kulturwissenschaft: Aby Warburg, die Pueblo-Indianer und das Nachleben der Antike', KUnstlcrischer AustauchArtistic Exchange. Alaen des XXV11. intemationalen KongressaJtlr Kunstgeschichu, dir. T. W. Gaehtgens (Akademie Verlig: Beriin, 1993), pp. 139-58. S. Weigel, 'Aby Warburgs Schlangenritual: Reading Culture and Reading Written Texts', New German Crltiaxie, vol. 22, no. 2, pp. 135-53; P.-A. Michaud, Aby Warburg a l'image en mourement (Macula: Paris, 1998), pp. 169-223, 247-80 (Warburg's unpublished notes for the 1923 conference). 5. Schlosser and Warburg were like-minded

The isolated and highly provisional experiment that I have undertaken here Is Intended as a plea for an extension of the methodological borders of our study of art (elner metlmdlschen Grenzerwelterung unserer Kunstwtssenschaft). . . . Until now, a lack of adequate general evolutionary categories has impeded art history In placing its materials at the disposal of the still unwritten - 'historical psychology of human expression' (hlstorischen Psychologic des menschllchen Ausdrucks). By adopting either an unduly materialistic or an unduly mystical stance, our young discipline blocks Its own panoramic view of history. It gropes toward an evolutionary theory of its own, somewhere between the schematisms of political history and the dogmatic faith In genius.1

Downloaded from http://oaj.oxfordjournals.org/ at UNAM on January 31, 2012

It was by approaching the image from an anthropological, then a psychological point of view that Aby Warburg was able to carry out the 'extension of methodological borders' that he defended before his colleagues at a conference in 1912. The immediate consequences of such an extension could only be disturbing for the discipline for it became clear that the time of the image is not the time of history in general, the time of the 'general evolutionary categories' that Warburg invokes here. What, then, is the most urgent task (as untimely and outdated today as it was in Warburg's epoch)? It is for art history to establish 'its own theory of evolution', its own theory of time. It is for art history to enter into a time other than habitual chronologies, eternal 'influences', old Vasarian or neo-Vasarian family myths.2 This other time is called 'survival' (Nachleben). The mysterious keyword or slogan of Warburg's entire enterprise, Nachleben der Antike, is by now familiar to us. It is the 'fundamental problem' which his archival research addressed, and for which he created the library that bears his name in order to grasp its sedimentations and shifting grounds.3 Warburg also confronted this 'fundamental problem' during the very brief period of his famous Native American experience.4 Therefore, before interrogating the notion of survival in the context of a 'science of culture' patiently worked out by Warburg using images from antiquity and the modern Western world it seems appropriate to situate the experimental emergence of this problematic on the limited and 'displaced' ground of his voyage to Hopi country. Anthropology's theoretical and heuristic function its capacity to de-territorialise fields of knowledge, to reintroduce difference in objects and anachronism in history will only appear in even sharper relief. The 'survival' that Warburg invoked and questioned throughout his entire lifetime is, above all, an Anglo-Saxon concept. In 1911, when Warburg's friend Julius von Schlosser referred to the 'survival' of figurative practices in wax, he did not rely upon the spontaneous vocabulary of his mother tongue.5 He did not write Nachleben, any more than he wrote Fortleben or Uberleben. He wrote survival, in English. This is significant evidence of a citation, a borrowing, a conceptual displacement, for what von Schlosser cites borrowed and displaced by Warburg before him is none other than the survival of the great British ethnologist Edward B. Tylor. In his sudden departure from Europe to Mexico in 189S, Warburg was not making a journey to archetypes, as Fritz Saxl thought, but a journey to survivals. And his
I OXFORD UNIVERSITY PRESS OXFORD ART JOURNAL 25 I 2002 5 9 - 7 0

Georges Didi-Huberman

theoretical landmark was not James G. Frazer, as Saxl wrote, but Edward B. Tylor.7 As far as I am aware, Warburg's commentators have not paid close attention to this anthropological source. At best, they have only considered the differences between Warburg and Tylor; Ernst Gombrich for example, argued that Tylor's 'science of culture' could in no way appeal to a disciple of Burckhardt whose major preoccupation was Italian art. And yet, this 'science of culture' was omnipresent in the opening of Primitive Culture (published in London in 1871), a work of such importance that, at the end of the nineteenth century, ethnology was commonly referred to as 'Mr Tylor's Science'.9 Of course, Primitive Culture's immense notoriety does not guarantee its status as a theoretical source. The point of contact between Warburg's Kuhurmssenschaji and Tylor's 'science of culture' lies first in the establishment of a particular link between history and anthropology. Indeed, both projects sought to overcome the eternal opposition which Levi-Strauss would continue to criticise a century later10 between the evolutionary model required by history and the type of atemporality with which anthropology is often credited. Warburg opened the field of art history to anthropology, not simply in order to recognise new objects of study, but also in order to open time.11 For his part, Tylor intended to carry out a rigorously symmetrical operation. He began by affirming that the fundamental problem of any 'science of culture' is its 'development of culture', and that this development is not reducible to an evolutionary law formulated according to models used by the natural sciences.12 Only through a history, or even an archaeology, of culture can the ethnologist understand its meaning:13
In working to gain an Insight Into the general laws of Intellectual movement, there Is practical gain in being able to study them . . . among antiquarian relics of no Intense modem Interest

with regard to many aspects of Warburg'i problematic. See Georges Didi-Hubermin, 'Viscosites et nirvivancej. L'histoire de l'art a l'epreuve du materiau', Critique 104, DO. 611, 1998, pp 138-62. 6. Julius von Schlosser, Gachichu da PonrStbildnerel in Wachs Ela Yersuch (1911), cd. T. Medicus (Akademie Verlag: Berlin, 1993), p 10. Warburg is dted on pages 76, 81-2, 186, 194. 7. Fritz Saxl, 'Warburg's Visit to New Mexico,' p. 326. 8. E. H. Gombrich, Aby Warburg: An Intellectual Biography (University of Chicago Press- Chicago, 1970), p. 16. Kurt W Fonter, 'Aby Warburg: Hu Study of Ritual on Two Continents', October 77, 1996, p. 6. Foster ignores Tylor as far as Warburg's 'ethnological culture' is concerned 9. See M. Panoff, 'Tylor (Sir Edward Burnett), 1832-1917', in Dictlonnalre du Darwinism a de I'erolutlon III, dir. Patrick Tort ( P U F : Paris, 1996), p. 4363. 10. Claude Levi-Strauss, 'Histolre et ethnologic' (1949), Anthropologie structurale (Plon: Paris, 1958), pp. 3-33. 11. WUlibald Sauerlander, 'Pour la deliverance du passe: Aby Warburg, une biographic mullectuelle par E. H. Gombrich', Histolre de l'art, nos 5-6, 1989, pp. 67. Sauerlander tecs, wrongly, tius dialectic as a pure and simple dUemma 12. Edward Burnett Tylor, Primitive Culture, researches into the derelopment of Mythology, Philosophy, Religion, An and Custom I, (Murray: London, 1871), pp. 23-62.

Downloaded from http://oaj.oxfordjournals.org/ at UNAM on January 31, 2012

Warburg certainly did not disavow this methodological principle of untimeliness: what makes sense in a culture is often the symptom, the un13. Tylor, Primitive Culture I, p. 13 thought, the anachronic aspect of this culture. Here, we are already within the 14. Tylor, Primitive Culture I, p. 143. spectral time of survivals. At the beginning of Primitive Culture, Tylor 15. Tylor, pp. 14-16. See E. B. Tylor, introduces this time theoretically by noting that the two competing models for Anthropology An Introduction to the Study of Man the 'development of culture' the 'theory of progress' and the 'theory of and Civilization (Macmillan: London, 1881), degeneration' must be thought dialectically, intertwined with one another. pp. 373400, where Tylor investigates the notions of 'tradition' and 'diffusion' The first The result would be a time knot difficult to untangle because evolutionary definition of 'survival' was offered by Tylor in movements, and movements that resist evolution, cross incessantly within it.15 1865: 'the "standing over" (superstitio) of old Through these crossings the concept of survival appears as a differential habits into die midst of a new changed state of dungs'. In Tylor, Researches into the Early History between two contradictory temporal states. of Mankind and the Derelopment of Qriiization Tylor dedicated an essential part of his work to the theoretical foundation of (Murray: London, 1865), p. 218 the concept of survival. But he had written die word, as if spontaneously, in 16. Tylor, Aaahuac. Or Mexico and the Mexicans, anodier context, in anodier temporality of experience, a displacement a trip Ancient and Modem (Green, Longman and Roberts- London, 1861), pp. 330-4. An index, to Mexico, to be precise. Between March and June 1856, Tylor crossed in two columns, is provided for all of diese Mexico on horseback, observing and taking diousands of notes. In 1861 he subjects. published his journal from the trip his own version of Tristes Tropiques in which, as if to his great surprise, mosquitoes and pirates, alligators and missionaries, die slave trade and Aztec relics, Baroque churches and Indian customs, earthquakes and the use of firearms, table manners and modes of account keeping, museum objects and street fighting, all enter the scene one after the odier. Anahuac is a fascinating book because we witness die audior's astonishment diat diis very experience, in this very place and moment, could bear such a knot of anachronisms, such a mixture of things past and present.
62 OXFORD ART JOURNAL 25.1 2002

The Surviving Image: Aby Warburg and Tylorian Anthropology

17. Tylor, Aoahvac, pp 47-54, 85-9. This anachronism is particularly evident in the system of illustrations, see especially, pp. 110-11, 220-1, 236. 18. Tylor, Primitive Culture I, p. 16. 19. See in particular the work of Tylor'J famous contemporary Gottfried Semper, Der Stil in da uchaachen and uktonhxhtn KUnsten. oder praklxhe Asthali- Bin Handbuch fir Technlktr, KUnaler trod Kunstfreunde (Bruckman: Munich, 1878-9) 20. Tylor, Primitive Culture I, p. 64. 21. Tylor, Primitive Cuhure I, p 63.

Thus, Holy Week festivities in Mexico bring heterogeneous, half-Christian, half-pagan commemorations up to date; the Indian market in Grande actualises a numbering system which Tylor thought could only be found in preColumbian manuscripts, the ornamentation of antique sacrificial knives was comparable with the spurs of Mexican Yaqueros.11 In the face of all of this, Tylor discovered the extreme variety, the breathtaking complexity of cultural facts (something one also notes in reading Frazer). Yet, he also discovered something even more overwhelming (which one never notes in reading Frazer): the vertiginous play of time in the present, in the present 'surface' of a given culture. Vertigo is first expressed in the powerful sensation in itself obvious, but its consequences less so that the present is woven with multiple pasts. This is why Tylor insists that the ethnologist must assume the historian's role in each of his observations. The 'horizontal' complexity of what he sees stems above all from a paradigmatic 'vertical' complexity of time:
Downloaded from http://oaj.oxfordjournals.org/ at UNAM on January 31, 2012
Progress, degradation, survival, revival, modification, are all modes of the connexion that binds together the complex network of civilisation. It needs but a glance Into the trivial details of our own daily life to set us thinking how far we are really Its originators, and how far but the transmitters and modifiers of the results of long past ages. Looking round the rooms we live In, we may try here to see how far he who only knows his own time can be capable of rightly comprehending even that. Here Is the honeysuckle of Assyria, there the fleur-de-lis of Anjou, a cornice with a Greek border runs round the celling, the style of Louis XIV and Its parent the Renaissance share the looking glass between them. Transformed, shifted, or mutilated, such elements of art still carry their history plainly stamped on them; and If the history yet farther behind Is less easy to read, we are not to say that because we cannot clearly discern It there Is therefore no history there.18

It is characteristic of this example of survival one of the first in Primitive Culture that it concerns the formal elements of ornamentation, the 'primitive words' of every notion of style.1 That this survival of forms is signified as a 'stamp' is equally distinctive. Admitting that the present bears the mark of multiple pasts means, above all, to allow for the indestructibility of an imprint of time, or times, on the forms proper to our present life. Therefore, Tylor speaks of 'the strength of these survivals' by which, using another metaphor, 'old habits maintain their roots in a ground overwhelmed by a new culture'. He also compares the strength of survival to a 'river which, having dug its bed, will run for centuries'. This is a way of elucidating always via the stamp what he referred to as the 'permanence of culture'.21 Warburg would have recognised his own investigation of permanence the tenacity of antique forms in the long duration of Western art history in this expression of a 'fundamental problem'. But that is not all. Such permanence could have been expressed, as it was in certain strains of nineteenth-century philosophical anthropology, in terms of an 'essence of culture'. The major interest of Tylor's thinking on this point, as well as its proximity to Warburg's approach, stems from this critical supplement: the 'permanence of culture' does not express itself as an essence, as a global feature or archetype, but on the contrary, as a symptom, as an exceptional feature, as a displaced thing. The strength of survivals, their 'power' even, as Tylor notes, is revealed in the tenuousness of minuscule, superfluous, derisory, or abnormal things. Survival, in itself, lies in the recurring symptom and in the game, in the pathology of language and in the unconsciousness of forms. So, Tylor turned his attention to children's games (bows and arrows, slingshots, rattles, knucklebones, or playing cards: survivals of the old and very serious practices of war and divination), just as Warburg would later turn his attention to Renaissance
OXFORD ART JOURNAL 25.1 2002 63

Georges Didi-Huberman

celebration practices. Tylor studied features of language, sayings, proverbs, and modes of salutation, just as Warburg hoped to do later for Florentine civilisation.22 Yet, Tylor was most specifically interested in the aspect of survivals which related to superstitions. He inferred the very definition of the anthropological concept of survival from its traditional, Latin meaning, superstnio:
Such a proceeding as this would be usually, and not improperly, described as a superstition; and, Indeed, this name would be given to a large proportion of survivals generally. The very word 'superstition', In what Is perhaps its original sense of a 'standing over' from old times, itself expresses a survival. But the term superstition now implies a reproach. . . . For the ethnographer's purpose, at any rate, It is desirable to Introduce such a term as 'survival', simply to denote the historical fact which the word 'superstition' Is now spoiled for expressing.23

22. Tylor, Primitive Culture I, pp. 63-100 23 Tylor, Primitive Culture I, pp. 645. 24. Tylor, Primitive Culture I, pp 101-44. 25. Aby Warburg, 'Italian Art and International Astrology in the Palazzo Schifanoia', and 'PaganAntique Prophecy in Wordj and Images in the
Age of Luther', in The Renewal of Pagan

Antiquity, pp 563-96 and pp. 597-697 respectively. 26. Tylor, Primitive Culture II, pp. 1-327 See
J. Pascher, Der Seelenbegrlff tm Ammismus E B. Tylors Eln Baiag zur Rellgionswissenschafi (Becker:

We can now understand why the analysis of survivals in Primitive Culture culminates with a long chapter dedicated to magic, astrology, and all of their related forms.24 How can we not recall the apex of the Nachleben der Antike, Warburg's analysis of the treatment of astrology in the Ferrara frescoes, or in Martin Luther's writings?25 In both cases and each time (and without even mentioning Freud), it is the flaw in consciousness, the fault in logic, the lack of sense in the argumentation which opens a breach, the breach of survivals, into the currency of a historical fact. Before Warburg and Freud, Tylor admired the capacity of 'trivial details' to make sense, or rather, be symptoms (which he also referred to as landmarks) of their own insignificance. Before Warburg and his interest in the 'animism' of votive effigies, Tylor, among others, attempted to construct a general theory of the power of signs.2 Before Warburg and his fascination for the expressive phenomena of the gesture, Tylor, again among others, attempted to construct a theory of emotional and mimetic language. Before Warburg and Freud, he staked out, in his own way, the lesson of the symptom absurdity, lapsus, sickness, madness as the privileged mode of access to the vertiginous time of survivals. Could the path to the symptom then, be the best way to hear the voices of ghosts?
Throughout the whole of this varied Investigation, whether of the dwindling survival of old culture, or of Its bursting forth afresh in active revival, It may perhaps be complained that Its Illustrations should be so much among things worn out, worthless, frivolous, or even bad with downright harmful folly. In fact It Is so, and I have taken up this course of argument with full knowledge and intent. For Indeed, we have in such enquiries continual reason to be thankful for fools. It Is quite wonderful, even if we hardly go below the surface of the subject, to see how large a share stupidity and unpractical conservatism and dogged superstition have had In preserving for us traces of the history of our race, which practical utilitarianism would have remorselessly swept away.28

Wurzburg, 1929). 27. Tylor, Primitive Culture I, pp 145-217. 28 Tylor, Primitive Culture I, pp 142. On this notion of conservatism see Tylor, 'ConservatismVariationInvention,1 (1874),
77K Collected Work of Edward Burnett Tylor VII

Downloaded from http://oaj.oxfordjournals.org/ at UNAM on January 31, 2012

(Routledge-Thoemmes Press- London, 1994), pp. 1378 (original pagination). 29 'On Traces of the Early Mental Condition of Man,' (1869) and 'On the Survival of Savage Thought in Modem Civilization,' (1869) in
Proceedings of the Royal Institution of Great Britain,

vol. 5 (1866-1869), pp. 83-93 and pp. 522-35 respectively See Margaret T Hodgen, 77K
Doctnne of Survivals. A Chapter in the History of Scientific Method in the Study of Man (Allenson:

London, 1936), pp. 67-107, 122, 130, 142.

Between phantom and symptom, the notion of survival becomes a specific expression of the 'trace' for the historical and anthropological sciences. As is well known, Warburg was interested in the vestiges of classical antiquity, vestiges which were in no way reducible to the existence of material objects, but could equally live on in forms, styles, behaviours, the psyche. We can easily understand his interest in Tylor's survivals. First, they marked out a negative reality, that which appears as a cast-off, ageless, out-of-date, or outof-use in a culture (just as, in the fifteenth century, Florentine boti testify to a practice already removed from the present and die 'modem' preoccupations on which Renaissance art was based). Second, Tylor's survivals marked out a masked reality; something persists and testifies to a vanished moment of society, but its very persistence is accompanied by an essential modification a change of its status of signification (to say that the bow and arrow of ancient wars have
64 OXFORD ART JOURNAL 25.1 2002

The Surviving Image: Aby Warburg and Tylorian Anthropology

survived as a children's game is to demonstrate the transformation of their


o

30. Aby Warburg, 'Pagan-Antique Prophecy', in 77K Renewal of Pagan Antiquity. 31. Sec, for example, S. A Cook, "The Evolution and Survival of Primitive Thought,' Essays and Studies Presented to William Ridgeway, dir. E. C. Quiggan (Cambridge University Press: Cambridge, 1913), pp. 375-^12; Arthur Weigall, Survirances paiennes dans le monde chtttlen (1928), trans. Ariane Flourny (Payof Pans, 1934); P. Saintyves, D marge de la Legende doree. songes, miracles et surrivances Essai sur la formation de quelques themes hagiographiques (Nourry- Pans, 1930); G. J. Laing, Surrivals of Roman Religion (Harrap: London, 1931) Chi the critical fortune of Tylonan survival, see M. T. Hodgen, 77K Doctrine of Surrirals, pp 10839, which unfortunately only treats the Anglo-Saxon domain. 32. See, for example, L C G Clarke, 'Modem Survivals of the Sumenan Chatelaine,' Essays Presented to C G Scligman, dir Edward Evan Evans-Pntchard et al (Kegan Paul, Trench & Trubner- London, 1934), pp. 41-7. Andre Leroi-Gourhan, Evolution et techniques I- L'homme a la matiere (Albin Michel. Pans, 1943), pp. 9 11 3 (on the notion of the 'technical stereotype'). 33 See M T Hodgen, 77K Doctrine of Surrirals, pp. 140-74. 34 Marcel Mauss, 'Essai sur le don Forme et raison de l'echange dans les societes archaiques' (1923-4), Scciologie et anthropologle (P U F.: Paris, 19S0), pp 228-57. 35. Mauss, 'Essai sur le don', p 228. 36 Mauss, 'La Volkskunde comme science' (1903), Oeuvres, 111 Cohesion sociaJc et divisions de la sociology, ed. Viktor Karady (Minuit- Paris, 1969), p. 372

status and their signification). In this sense, the analysis of survivals seems to be the analysis.of symptomatic and ghostly manifestations. They designate a reality of effraction, if tenuous, or even imperceptible, and for this reason they also designate a spectral reality. Astrological survival will thus appear as a 'ghost' in Luther's discourse, a ghost whose effectiveness Warburg recognised because of its intrusive and interfering nature as symptom in the logic of the preacher of the Reformation preacher's argumentation. It comes as no surprise that the critical richness of Tylor's survivals first concerned phenomena of belief: the first applications of this concept took place in the domain of the history of religions.31 Nevertheless, in anticipation of what Andre Leroi-Gourhan called 'technical stereotypes', several archaeological studies of long durations approached the history of objects from the angle of survival.32
Downloaded from http://oaj.oxfordjournals.org/ at UNAM on January 31, 2012

We must note, however, that the notion of survival has never been very well received and not only by art history. In Tylor's time, survival was accused of being too structural and abstract a concept, a concept which defied all precision and factual verification. The positivist objection consisted in asking: but, how do you date a survival? This is precisely to misunderstand a concept that meant to identify a non-'historical' in the trivial and factual sense type of temporality. Today, one would accuse survival of lacking structure, of being a concept, as it were, marked by the evolutionist seal, therefore out-of-date, and outdated; in short, an old nineteenth-century scientific ghost. One might spontaneously infer this from modern anthropology which, from Marcel Mauss to Claude Levi-Strauss, effected the necessary reorientation of too essentialist (Frazer) or too empiricist (Malinowski) ethnological concepts. However, in bringing out the critical aspects themselves, one notices that things are more complex and nuanced than they appear. Survival itself is not in question, but a certain use-value made of it by several Anglo-Saxon ethnographers during the nineteenth century. Mauss, for example, did not hesitate to use the term. The third chapter of Essai sur le don (Essay on the Gift) is titled 'Survival of these principles [where an "exchange of gifts" is carried out] in ancient rights and economies.' There he explained that the principles of the gift and counter-offer count as 'survivals' for the historian and for the ethnologist:
They have a general sociological value, since they allow us to understand a moment of social evolution. But there Is more They also have an Importance for social history. Institutions of this type have effectively provided the transition to our forms, our own forms, of law and economy. They serve as historical explanations of our own societies. Morality and exchange practices In use by the societies which Immediately preceded our own maintain the more or less Important traces of all of the principles Just analysed [In the framework of so-called primitive societies].35

Elsewhere, Mauss went so far as to extend the notion of survival to 'primitive' societies themselves:
There Is no known society which has not evolved. The most primitive of men have an Immense past behind them; thus diffuse traditions, and survival play a role even for them.3*

This was not only a manner of saying that 'primitive societies have a history' which some had long denied, hence the expression 'people without history' but that this history may be as complex as our own-. It, too, is made up of the conscious handing down of 'diffuse traditions', as Mauss wrote. It, too, forms
OXFORD ART JOURNAL 25.1 2002 6S

Georges Didi-Huberman

itself within a play of or a knot of heterogeneous temporalities a knot of anachronisms, even if the absence of written archives makes this difficult to analyse. Mauss does not criticise the usage of survival in order to call into question this complexity of temporal models. On the contrary, he does so to refute ethnological evolutionism as a simplification of temporal models. Thus, when Frazer describes survival as a 'confusion between ancient magic and religion', Mauss responds that 'the hypothesis does not tell us very much'; that is, the hypothesis that there was a confusion between magic and religion followed by the autonomisation of the latter, which became more rational and moral, in sum, more 'evolved'.37 Mauss also perceptively criticised what we might call archetypism, the remaining key trap in any analysis of survivals. Archetypism not only resulted in the simplification of temporal models, it led to their pure and simple negation, their dilution into an essentialism of culture and the psyche. The principle lure of such a trap is analogical perception. When resemblances turn into pseudomorphisms; when they serve to bring out a general and atemporal signification, then survival of course becomes a mythification, an epistemological obstacle. It must be pointed out that it has been possible to interpret and use Warburg's Nachleben to such ends. However, what distances Nachleben from any such essentialism is Warburg's philological effort, his perception of singularities, his constant attempt to tug at all direads, to identify each strand even when he knew that the threads escaped him, had been broken, or ran underground. Symptomatic anamnesis has strictly nothing to do widi archetypal generalisation. Levi-Strauss's criticism in the introductory chapter of Structural Anthropology seems even more severe. Because it is more radical, it is more partial, and sometimes inaccurate, if not disingenuous. Levi-Strauss starts off by walking in Mauss's footsteps: he criticises archetypism and its erroneous usage of substantialised analogies, of universally applicable pseudomorphisms. Now, it turns out that he sought its traces in Tylor's work. The bow and arrow no longer form a 'species', as Tylor had claimed in a language modelled on die biological bond of reproduction, because 'between two identical tools, or between two different tools which are as alike in form as they can be, there is and there will always be a radical discontinuity, which comes from the fact that one is not issued from another, but each of diem are issued from a system of representation'. Note that Warburg would have agreed widiout hesitation to this first point, which situated the organisation of symbols as the founding structure of the empirical world. Levi-Strauss stumbles when he goes one step further, claiming diat studies stemming from a problematic of survivals 'teach us nothing about unconscious processes translated into concrete experiences'. He invalidates this a few pages later by granting Tylor a nearly fundamental place in the evaluation of the 'unconscious nature of collective phenomena'. But to his mind, Tylor's ethnology remained devoid of any historical character. He cites as evidence a
brief passage in Researches into the Early History of Mankind (1865), without even

37. Mauss, 'La magic selon Frazer,' (1913), Orurra, I. Lafonaions soaala du sccrt , ed. V. Karady (Muiuit: Paris, 1968), p. 155. 38. 'Tous lea faita de sjroiljtude ne sont pas des faits de "recurrence", d'invenrion independante, de "survivances" des souches devolutions parfaitement ldenOque partout Maij invenement touj ne ont pas dej fait d'emprunts, et surtout pas des emprunts a un seul foyer. [. .] Nous demandons done qu'on mette, dans toute cette soi-disant histoire soaologique, moins de soaologie et plus d'histoire, et que chaque phenomene soit appreae en lui-meme.' Mauss, 'La theone de la diffusion urucentrique de la civilisation,1 (1925), Oeurra, II Representations collecttra a direntus da drillsatlons, ed. V Karady (Minuit: Paris, 1974), pp. 522-3.

Downloaded from http://oaj.oxfordjournals.org/ at UNAM on January 31, 2012

39. Claude Levi-Strauss, 'Histoire et ethnologic', Anthmpologie urucuiralc, p. 6. 40. Levi-Strauss, 'Histoire et cthnologie', p. 7 41. Levi-Strauss, 'Histoire et ethnologic1, pp. 9, 25. 42. Levi-Strauss, 'Histoire et ethnologic', pp.7, 13-14. 43. 'Un peuple pnmitif n'est pas davantage un peuple sans histoire, bien que le deroulement de celle-d nous echappe souvent [. . .] l'hlstoire de ces peuple nous cst totalemcnt inconnue, et, en raison de l'absence ou de la pauvrete des tradition oraleset des vestiges archeologiques, elle est a jamais hors d'atteinte On n'en saurait conclure qu'elle n'existe pas.' Levi-Strauss, 'La Notion de l'archaisme en ethnologic',
Anthropologie structuralc, p p . 11415.

taking into account the title of the work. Moreover, Levi-Strauss does not recognise that, six years later, in Primitive Culture, Tylor developed a reflection on the historicity of primitive cultures that he credits only to Franz Boas.42 In 1952, die author of Structural Anthropology announced a thesis on die 'unattainable' historicity of primitive peoples, clearly an entirely unconscious paraphrase of the passages from Tylor cited above.43 None of this changes the fundamental question, which is still a matter of knowing what survival means. And it is first a matter of knowing how, in what
66 OXFORD ART JOURNAL 25.1 2002

The Surviving Image: Aby Warburg and Tylorian Anthropology

44. Tylor, Researches into the Early History of Mankind, pp. 150-90. 45. Sec M. T. Hodgcn, The Doctrine of Survivals, pp 36-66. 46 'It may have struck some readers as an omission that in a work on civilization insisting so strenuously on a theory of development or evolution, mention should scarcely have been made of Mr Darwin and Mr Herbert Spencer, whose influence on the whole course of modern thought on such subjects should not be left without formal recognition. This absence of particular reference is accounted for by the present work, arranged on its own lines, coming scarcely into contact of detail with the previous work of these eminent philosophers.' Tylor, Primitive Culture, pp. VII-VIII. The global link that Mauss established between the 'English Anthropological School' and Spencerian evolutionism deserves clarification. See Marcel Mauss, 'L'ccolc anthropologique anglaise et la theorie de la religion selon Jevons' (1898), Oeuvres I, pp. 10910. Tylor's fundamental references at the beginning of Primitive Culture, m fact, belong to the German School of Anthropology: A. Bastian, Mensch in der Gerschtchte Zur Begriindung einer psychologlschen Westanschauung (Otto Wigand. Leipzig, 1860); T. Waitz, Anthropologie der Naturrolker (Fleischer Leipzig, 1860-72) On these problems, see Robert H. Lowie, 77K History of Ethnological Theory (Holt, Reinhart and Winston: New York, 1937), pp 68-85, and above all, J. Leopold, Culture In Comparative and Evolutionary Perspective: E. B. Tylor and the Making of 'Primitive Culture' (Dietrich Reimer: Berlin, 1980) 47. See M. T. Hodgen, The Doctrine of Survivals, p. 40, and especially J Leopold, Culture in Comparative and Evolutionary Perspective, pp. 49 SO, which demonstrates the complexity of the epirtemological sources of Tylorian survival. 48 See M. Panoff, 'Tylor', pp 4364-5. 49. E. H. Gombrich, Aby Warburg, p. 68,168, 185, 321, etc. and 'Aby Warburg e l'evoluzionismo ottocentcsco', Belfagor, 159, no. 6, 1999, pp. 635-49. 50. E. H. Gombrich, 'Aby Warburg e l'evoluzionismo ottecentesco', pp 6357, 645. 51. 'Cette nouvelle dimension [qu'apportait Darwin dans les sciences de la vie] e'etait le temp* et l'histoire. Sans doute Lamrack avait anteneurement accorde a l'immensite de la duree cosmique le pouvoir de produire fuccessivement a l'exutence la serie continue et progressive, quoique parfois lrreguhere, des corps organises, "depuls le plus imparfalts jusqu'aux plus parfaits." Mais la nouveaute radicale de L'Orlgine des espices consistait enceci qui le temps de lavie n'y etait pas suppose comme un pouvoir, mais qu'il etait percu directement dans de effeu en apparence distincts, en reahte unifies par leur complementarite. Le fossile, e'etait le temps petrifie, l'embryon, e'etait le temps operant;

sense, and for what stakes, this concept takes over or does not take over evolutionist doctrine. When Tylor peppers the seventh chapter of his book Researches into the History of Mankind (devoted to the 'Development and Decline of Civilisation') with references to Darwin, the stake is clearly polemical. In this text, he has to play human evolution against divine destiny; that is, he plays The Origin of Species against the Bible.44 He had to rehabilitate 'developmentalism' and the perspective of the species against religious theories of degeneration and the perspective of original sin.45 One further element needs to be made clear. The vocabulary of survival had not yet been set out when Tylor entered into this reference game. Even if the debate over evolution constitutes his general epistemological horizon, Tylor's notion of survival would be clearly constructed independently of Spencer's and Darwin's doctrines. Whereas natural selection referred to the 'survival of the fittest', the guarantor of biological innovation, Tylor approached survival from the opposite angle, from the angle of the most 'unfit, or inappropriate' carriers of a bygone past, instead of an evolutionary future.47 In short, survivals are only symptoms that carry temporal disorientation. They have nothing whatsoever to do with the premises of a teleology in progress, or with any 'evolutionary sense'. They certainly bear the evidence of a more original and repressed state, but they say nothing about evolution in itself. They undoubtedly have a diagnostic value, but no prognostic value. It is important to recall that Tylor's theory of culture stemmed neither from a biology nor from a theology. For him, 'savages' were no more the fossils of an original humanity than degenerates from a likeness to God. Instead, his theory aimed at a historical and philological perspective, which is why Warburg would take such an interest in it.4 One thing is certain: Warburg's concept of survival (Nachleben) was first sketched out within an epistemological field bound to anthropological objects, and toward the general horizon of evolutionary theories. In this sense, as Gombrich affirms, Warburg remained a man of the nineteenth century. In this sense, his history of art remains old-fashioned, its fundamental theoretical models outdated.49 The simplification is brutal, and not devoid of bad faith. At best, it demonstrates the difficulty that second generation iconologists faced when coming to terms with a legacy that was far too ghostly to be 'applicable' as such. At worst, this simplification aims to close off the theoretical paths opened by the very notion of Nachleben. What does it mean that Warburg was an 'evolutionist'? That he read Darwin? Of this, there is not the shadow of a doubt. That he defended an 'idea of progress' in the arts and adopted a 'continuist model' of time? Nothing is further from the truth. Evolutionary doctrine introduced the question of time into the life sciences beyond the 'long cosmic duration' in the words of Georges Canguilhem that had framed Lamarck's thought. However, posing the question of time already meant posing the question of times, of the different temporal modalities that make up, for example, a fossil, an embryo, or a rudimentary organ.51 Furthermore, Patrick Tort has shown the abusive error involved in reducing Herbert Spencer's philosophy which one recalls immediately whenever one hears the word 'evolutionism' to the Darwinian theory of biological evolution. The latter is a bio-ecological transformation of the development of living species inasmuch as they are subject to variation. The former is a doctrine, an ideology whose conclusions circulating amongst the nineteenthcentury ruling classes and industrial milieus are opposed to many of the
points made in The Origin of Species.
OXFORD ART JOURNAL 25 1 2002 67

Downloaded from http://oaj.oxfordjournals.org/ at UNAM on January 31, 2012

Georges Didi-Huberman

The misunderstanding pivots precisely around the notion of survival. Only in the fifth edition did Darwin insert the Spencerian expression of the 'survival of the fittest'. Today, epistemologists see nothing but theoretical confusion in the association of these two words (which Tylor, as we have seen, carefully dissociated). Speaking in such a way effectively reduces survival to selection. The most adept, the strongest, survive and reproduce. The idea that this law could be of relevance to the historical or cultural world comes from Spencer, not from Darwin, who instead saw in civilisation a means of opposing of 'disadapting' from natural selection. In this sense Warburg was undoubtedly a Darwinian, but not a Spencerian, evolutionist. For Warburg, Nachleben meant making historical time more complex, recognising specific, non-natural temporalities in the cultural world. Basing a history of art on 'natural selection' through the successive elimination of the weakest styles, thus providing evolution with its perfectibility and history with its teleology is in opposition to his fundamental project and his temporal models. For Warburg, the surviving form does not triumphantly outlive the death of its competitors. On the contrary, it symptomatically and phantomatically survives its own death: disappearing from a point in history, reappearing much later at a moment when it is perhaps no longer expected, and consequently having survived in the still poorly defined reaches of a 'collective memory'. Nothing is further from this idea than Spencer's 'synthetic' and authoritarian systematism, his so-called 'social Darwinism'.54 On the other hand, links can be traced between this notion of survival and certain of Darwin's terms relating to the complexity and paradoxical intricacy of biological time. From this perspective, Nachleben could be compared with, but not assimilated to, temporal models that precisely create symptoms in evolution; that is, models that set up obstacles within all continuity-based adaptation schemes. Evolutionary theoreticians have spoken of 'living fossils', perfectly anachronistic beings of survival.55 They have spoken of 'missing links', intermediary forms between ancient and more recent forms of variation.56 With the concept of 'retrogression', they have refused to oppose a 'positive' evolution and a 'negative' regression.57 They have not only spoken of 'panchronic forms' Living fossils or surviving forms, organisms that were believed to have disappeared, or that had been found everywhere fossilised, which were suddenly discovered as living organisms under certain conditions58 but also of 'heterochronies', paradoxical states of living which combine heterogeneous phases of development.59 At those moments when the usual game of natural selection and genetic mutation does not enable the understanding of a new species, they have even spoken of 'promising monsters', 'non-competitive organisms', nevertheless capable of engendering an original, radically divergent evolutionary line.60 Indeed, in its own way, Warburgian Nachleben only tells us about 'living fossils' and 'retrogressive' forms. It tells us about 'heterochronies', and even about 'promising monsters' like Durer's Landser sow, with her two bodies and eight hooves, which Warburg treated from the perspective of what he referred to as a 'region of prophetic monsters' (Region der wahrsagenden Monstra).61 It is easy to understand that a work as experimental, as disquieting, and as heuristic as Warburg's could be misunderstood as 'evolutionist'. In order to discern the anachronistic and unprecedented object of his quest, Warburg forged ahead like all pioneers. He assembled a system of heterogeneous debts, whose orientation could be changed by simply comparing them with all of the others. What other conclusion should we
68 OXFORD ART JOURNAL 25.1 2002

l'organc rudimentaire, c'etait lc temps retarde [ . .] La classification ccssait d'etre une peinture des formej coexistantes pour devenir un canvas synopaque tisse avee les fils du temps ' Georges Canguilhem, ldeolgle et
TQUonahtk dans l'hlstolre da sciences de la ne Nourelles etudes d'hlstotre a de philosophie des

sciences (Vrin: Paris, 1997), p. 106. 52. 'Une enorme erreur de methodc ct un contresens dieorique d'unc extreme envergure regnet encore sur 1'apprehension globale de la theone darwinnienne. Pendant pluj d'un siede, sous 1'influence d'un contrat enondatif passe entre l'ascension de 1'industnahsme liberal anglais et de la philosophie synthetique de Spencer, la theone de la selection naturelle a servit de garantie et de modele scientiiique a des doctrines, a des recommandations et a des pratiques soaales et pohtiqucs qui se faisaient passer pour les consequences directement apphcatives de la loi nucleaire dc 1'evohition biologique- celle de la competition et de l'exclusion ehminatoire ' Patrick Tort, La pensee
hitrarchique et 1'erolution (Aubner Montaigne

Downloaded from http://oaj.oxfordjournals.org/ at UNAM on January 31, 2012

Pans, 1983), pp. 166-97 and as already ated (which opposes the scientific paths of biological evolution and the ideological derailment of Spenccnan evolution) and 'L'effet reversif de revolution. Fondements de l'anthropologie darwinienne', Darwinlsmc et socicte, dir. Patrick Tort (P U F : Pans, 1992), pp. 1 3 ^ 6 . 53. 'La selection naturelle selectionne la civilisation, qui s'oppose a la selection naturelle ' Tort, 'L'effet reversif de revolution', p. 13. See D Becquement, 'Survivancc du plus
aptc', Dictionnane du darwinteme et de 1'erolution

III, dir. Patrick Tort (P.U.F Pans, 1996), pp 4173-5.


54 See H Spencer, The Study ofSociology

(1873), ed. T. Parsons (University of Michigan Press: Ann Arbor, 1961). 55. See N. Eldredge and S. M Stanley, hnng Fossils (Springer. New York, 1984). 56 See C Devillen, 'Formes Intermediares
(chainons manquants)', Diatonnatre du darwintsme et de 1'erolution II, pp. 15947. 5 7 . See P. T o r t , 'Evolution regressive,' Diaionnaire du darwtnisme I, 'On parle d'erolution

rearessire pour caractenser la regression qui frappe certains organes devenui inutiles ou nuisibles a l'espece. L'exemple donne par Darwin est celui des insectes insulaircs continueUement exposes. L'exemple donne par Darwin est celui des insectes insulaires continueUement exposes a la force des vents, et Cjiri perdent leurs ailes (p. 1595).
'Retrogression', Diaionnaire du djjnrinisme III,

p 3677. 58. See M. Delsol and J Flatin, 'Formes


panchroniques', Diaionnaiic du danrinisme II,

p. 1714-17. 59. See C. Devillers, 'Heterochronies,'


Diaionnaire du danrinisme II, pp. 221517, which

The Surviving Image: Aby Warburg and Tylorian Anthropology

give* as an example the Axolotl (which 13 both infant and adult and remains a larvae capable of reproduction throughout its entire life) and speaks of differential rhythms, developmental acceleration, or slowing down, known as "neotenies", progeneses, peramorphoses, hypermorphoses, etc. See also K. J. McNamara, 'Heterochrony and Phylogenetic Trends', Palcohwlogy 8, 1982, pp. 130-42. 60. See M. Delsol, 'Monstres prometteurs',
Dtctlannalrt du darwinisme II, p p . 304244.

draw from this play of debts and debated questions if not that, there, evolutionism produced its own crisis, its own internal critique? By recognising the need to broaden canonical models of history narrative models, models of temporal continuity, models of objective realisation by directing himself little by little toward a theory of the memory of forms made up of leaps and latencies, Aby Warburg decisively broke with notions of 'progress' and historical 'development'. He thus played evolutionism off against itself. He deconstructed it solely in order to recognise phenomena of survival and cases of Nachleben which now must be dealt with in terms of their specific development.
Translated from the French by Di Vivian Sky Rchberg.

61 Warburg, 'Pagan-Antique Prophecy in Words and Images in dSe Age of Luther', p. 635 62. A. Dal Lago, 'L'arcaico e il suo doppio', Am out, nos. 199-200, 1984, pp 79-86. See also Leopold D. Etthngcr, 'Kunstgeschichte als Geschichte', (1971-76), AusgewShhc Schnjicn, pp. 499-513

Downloaded from http://oaj.oxfordjournals.org/ at UNAM on January 31, 2012

OXFORD ART JOURNAL 25 I 2002 69